Skip to contents

Simulation of typical subjects

Source: vignettes/NMsim-TypSubj.Rmd
NMsim-TypSubj.Rmd

Objectives

This vignettes aims at enabling you to use NMsim for the following purposes

  • Simulation of typical subjects

Simulation of a typical subject

A typical subject is here understood as a subject without random effects, i.e. all ETA’s equal zero. It is important to realize that “typical” does not have to do with covariates which the user will still need to control in the model, in the simulation input data, or by a combination of these. Getting NMsim to run with all ETA’s equaling zero is this easy:

simres.typ <- NMsim(file.mod=file.mod,
                    data=dat.sim,
                    name.sim="typSubj",
                    typical=TRUE)

In the first simulation we used PRED from the default simulation method to get a typical subject simulation. That will work in many cases, but that depends on the model. The way to run a simulation with all ETA’s set to 0 is using method.sim=NMsim_typical.

p.typ <- ggplot(simres.typ,aes(TIME,IPRED,color=trt))+geom_line()+
    geom_line(aes(y=PRED))
p.typ